In communications networks, many methods and systems exist for optimizing the efficiency of data transfers throughout a data network. These approaches include using optical code division multiple access (“OCDMA”) data networks, switched interconnects, collision based interconnects, separate interconnects and common higher bandwidth protocol interconnects. However, each of these approaches has significant drawbacks. For example, with respect to native optical code division multiple access (“OCDMA”) data networks, a problem is found whereby when two or more encoders send data via a communication channel to the same decoder an undetectable data collision occurs. Thus, native OCDMA techniques do not support organized channel sharing by two or more encoders. Also, a switched interconnect approach requires a centralized switch for channel arbitration and bandwidth allocation. With a collision based interconnect approach, an undesirable latency period in determining bandwidth allocation based on detected data collisions is created. A separate interconnect approach requires separate physical interfaces to provide sufficient channel-to-channel isolation. A common higher bandwidth protocol interconnect approach does not allow for variable bandwidth allocation and requires all nodes in the data network to be converted to a single high bandwidth protocol.